What You Can Do With AI Right Now

⏳ Read time: under 12 minutes

Welcome back, Aioli enthusiast

It’s been a few weeks since I sent out my last edition of this newsletter— I’ve been busy tinkering with some new AI tools and building out AI-enhanced systems for small businesses.

If you were familiar with my brand a few months ago, you may notice I’ve switched up the look a bit. It’s a work in progress 🙂 I’ve been creating content for Youtube behind the scenes, too.

The AI landscape continues to change week after week and AI has gotten substantially smarter since the last time you read this newsletter. The tech rumor mill has been swirling with theories that OpenAI has had a reasoning breakthrough with their unreleased project, “Strawberry”, fueled by a seemingly innocent tweet by Sam Altman remarking on summer in the garden 🍓 The implications are a significant step towards AGI (Artificial General Intelligence) — a vaguely defined concept for an AI that outperforms humans at every task — which has the potential to reshape industries, accelerate scientific discovery, and fundamentally alter our relationship with technology. I’ll believe it when I see it.

New feature alert! You can now listen to this newsletter 🤯 Try it out by clicking “Listen Online” in the top right corner of this email👇️ 

Today’s Spread

• 🧵 10 reasons to start using Claude instead of ChatGPT

• 📸 Can you supplement your content strategy with AI-generated product photography? These examples might change your mind

• 🍣 Your sushi is now AI-optimized

• 📊 How AI is being misused (AKA what to look out for)

• 📰 This week’s AI news to know
  
  📬 _{Did someone forward you this email? Sign up to get this newsletter here}

AI Use Cases

🧵 10 Amazing Things You Can Do with Anthropic’s Claude Artifacts

Since Claude 3.5 Sonnet was released in June, I have been using it almost exclusively over ChatGPT or other available models. I find it to be a better creative writer and better at critical reasoning.

One of OpenAI’s very own co-founders left for Anthropic (the makers of Claude), if that says anything 👀 

If you haven’t tried Claude, make time to test it out - it’s free!

First, what are Claude Artifacts?

Artifacts allow Claude to share substantial, standalone content with you in a dedicated window separate from the main conversation. Artifacts make it easy to work with significant pieces of content that you may want to modify, build upon, or reference later.

Some common examples of Artifact content include:

• Documents (Markdown or Plain Text)

• Code snippets

• Websites (single page HTML)

• Scalable Vector Graphics (SVG) images

• Diagrams and flowcharts

Pro tip: You have to turn on Artifacts in Claude to be able to use it. To enable the Artifacts feature preview, click on your initials in the lower left corner in Claude and select “Feature Preview” from the dropdown menu. From the proceeding menu, turn the Artifacts toggle on.

Now let’s look at some examples of what it can do…

1. Create interactive dashboards (linked above)

2. Code up games: Someone whipped up a fully functional Mancala web app in just 25 seconds, using nothing but a screenshot of the game rules. (link)

3.  Simplify data visualization: A savvy analyst used AI to crunch numbers, create eye-catching visuals, and build simple presentations - all without leaving the chat window. (link) 

4. Customize personal calendars: Someone took on Google Calendar by building their own basic clone. (link)

5. Create a functional soundboard: One creative dev created an AI sound effects generator using the ElevenLabs API. (link) 

6. Build interactive dashboards: A pitch deck got a major upgrade when someone transformed it into an interactive dashboard for investors - all with a single prompt. (link) 

7. Generate infographics and SVGs: One user pushed AI to its limits by creating an infographic from a 78-page PDF and even recreating a painting as an SVG. (link) 

8. Animate physics simulators: Someone created a 3D physics simulation using WebGL in just one try. (link) 

9. Create decision-making bots: By feeding massive amounts of data and preferences to AI, one user developed an interactive decision-making bot. (link)

10. 3D space animation on demand: Someone created a complex 3D solar system animation based on a simple text prompt in under 60 seconds. (link)

Here’s a sample prompt to test out Claude Artifacts for yourself:

Create a website mockup for my company, [insert description and mission statement here]

Even if you already have a website, it can surely be improved. Claude will give you tangible feedback and layout improvements. I do this by attaching screenshots of landing pages and adding in a ton of context (visitor analytics, sales goals, target consumer information, etc.) Much cheaper than hiring a digital marketing consultant!

Let me know if you try out Claude Artifacts and if you have any exciting use cases to share.

AI for Marketing

📸 AI-Generated Product Photography?

As someone who's spent plenty of time behind the camera shooting products and lifestyle scenes for CPG brands, I've always been a bit doubtful about AI's ability to create photos good enough to replace actual photoshoots. But I will admit, my mind has started to change since stumbling across these examples (see below).

Don't get me wrong – I still don't think AI is going to completely take over the world of product photography. But after seeing what it can do, I've got to admit it could definitely add something to the mix.

These AI-generated images are pretty impressive, and they've got me thinking about how they could fit into the world of product shots for PDP images, ad creative, and potentially even social content.

Keep in mind, this requires a great deal of prompt engineering and refinement to get to this point!

Tools used: Midjourney (I have a tutorial on how to access this. Message me if you’re interested or need tips!)

AI for Operations

🍣 Your Fish is Grown with AI Insights

Next time you're at a sushi bar, consider this: that tuna on your plate might have had a better health data plan than you do. Canadian startup Wittaya Aqua is bringing cutting-edge AI to fish farms.

With global seafood consumption soaring, this tech is very timely. Did you know that 3 billion people around the globe rely on wild-caught and farmed seafood for their protein intake? Wittaya recently reeled in $2.8 million in funding to cast their net wider, particularly into Asia – the world's aquaculture powerhouse.

Traditionally, aquaculture data has been fragmented and slow-moving, restricting the ability to make well informed decisions for aquatic animal production.

Their AI platform crunches numbers on everything from optimal feed to growth predictions, helping farmers boost efficiency and sustainability.

Wittaya Aqua's platform blends data points, AI and machine learning to:

• Forecast animal growth

• Recommend optimal feed types and quantities

• Analyze historical data and environmental factors

• Suggest strategies for maximizing crop yields

This innovation serves as another reminder to look at your own data points and see what you might be overlooking to optimize your processes. How could AI help you make sense of it all?

I advise companies on how to leverage their data and AI for insights and forecasting. If you’re interested in learning more, reach out. 

AI Ethical Use & Responsibility

❌ Here’s How Generative AI is Being Misused

When discussing the dangers of AI technologies, I often share an alarming example: with just a few seconds of someone’s voice recording, I can create a convincing voice clone without their consent. While I would never do this, bad actors are exploiting these accessible tools to scam innocent people.

As artificial intelligence becomes more powerful and accessible, concerns about its potential misuse— like the above example— grow.

Though I am a huge proponent of using AI in the appropriate settings, I also want to make it clear that it is equally important to be mindful of its dangers and pitfalls. Hence, why I want to share both the good and the bad.

A recent study by Google Deep Mind sheds light on how multimodal generative AI is being exploited today, offering valuable insights for developing safer AI technologies (and what to be cautious of).

Here are the key findings:

• Common tactics: Impersonation, synthetic personas, evidence falsification

• Primary goals: Influence public opinion, enable scams, generate profit

• Observations:

  • Most cases use consumer-level AI tools

  • Traditional tactics amplified by AI

• Ethical concerns:

  • Undisclosed AI use in politics

  • AI-generated voices of deceased for advocacy

• Technical misuse: Removing safeguards, causing AI malfunctions

• Implications:

  • Easier creation of deceptive content

  • More convincing manipulation

  • Need for better safeguards and awareness

• Limitation: Based on potentially biased media reports

This research underscores the urgent need for improved AI safety measures and increased public awareness about AI's potential risks and responsible use!

Speaking of AI education and awareness, I have a quick favor to ask you.

I spend several hours crafting this email entirely for free (hence why I’ve taken a little break to focus on client work). It would mean the world to me if you could share Aioli with just 1 person who you think could benefit from learning about AI.

Here’s your unique referral code you can copy and paste:

Merch

🤠 Aioli Merch!

Want to rep an Aioli hat but don’t want to go through the trouble of referring 10 people? I listed it on my website for purchase (with free shipping!)

I’ve been wearing mine since mid-March. Fun fact: someone who had seen my posts even recognized me on a flight from Austin to SFO because of it! It’s been a good conversation starter 💬 

AI News

🍴 Quick Bites

🇰🇷  South Korea is developing "future classrooms" equipped with AI-powered textbooks. These textbooks can evaluate a student's reading level and adjust the content accordingly.

🫣 OpenAI has developed a tool that can reportedly identify text generated by ChatGPT with 99.9% accuracy. While this technology exists, the company has not made it publicly available.

🦈 Researchers at UC Santa Barbara have developed an AI model that can identify sharks in water with potentially greater accuracy than humans.

🏓 Google Deep Mind built a ping-pong playing robot that wins 45% of matches against humans with varying skill levels

Thanks for reading. Your curiosity fuels this journey.

Until next time!

Meg 👩🏻‍💻

p.s. if you want to sign up for this newsletter or share it with a friend or colleague, you can do so here. For questions and feedback, email me at [email protected]